# Thread: hierarchical regression for controlling variables?

1. ## hierarchical regression for controlling variables?

For my dissertation, my supervisor told me to do a regression analysis and control for 2 variables. So should I report in my dissertation that I have done a hierarchical regression analysis or should I report that I have done a standard multiple regression analysis while controlling for the 2 variables?

I thought reporting that I have done a hierarchical regression analysis is better, what do you think?

2. ## Re: hierarchical regression for controlling variables?

If all you did was a multiple regression and included 2 other variables you wanted to control for I don't see how you could call it a hierarchical regression.

3. ## Re: hierarchical regression for controlling variables?

I wil explain that more clear:

she told me to regres the independent variable (x) on the dependent variable (y), and by doing so control for the effects of 2 variables (clicking on "next").
So my question is whether I should report that I have done a hierarchical multiple regression or a standard multiple regression?

4. ## Re: hierarchical regression for controlling variables?

Still sounds like a standard multiple regression (with three DV, two of which were only in the model to control for). I don't think you know what a hierarchical model is, it is something in particular were you are controlling for structural levels. Might help if you list out your DVs.

5. ## Re: hierarchical regression for controlling variables?

It's still not clear to me what you've done exactly and why this would be hierarchical at all.

6. ## Re: hierarchical regression for controlling variables?

Just describe it in terms of which variables are or are not controlled for. That is the most straightforward and easy to understand description.

7. ## Re: hierarchical regression for controlling variables?

Dason: I guess it could be thought of as hierarchical in the sense that first covariate 1 was entered, then covariate 2 was entered, and then the predictor of interest was entered. I don't think this is a helpful way to think about it, but it would make it fit under the rubric of a hierarchical regression.

8. ## Re: hierarchical regression for controlling variables?

Oke

independent variable: stress
dependent variable: eating
control variables: gender, education and age

Thus: regressing stress on eating, while controlling for the effects of gender, education and age.
Note that I found that gender, education and age did not explain any variance in eating (model 1 was insignificant).

9. ## Re: hierarchical regression for controlling variables?

Originally Posted by Jake
Just describe it in terms of which variables are or are not controlled for. That is the most straightforward and easy to understand description.

YES! that is exactly what I mean

Can I report that I used hierarchical regression, then?

10. ## Re: hierarchical regression for controlling variables?

Like I said earlier, I wouldn't.

11. ## Re: hierarchical regression for controlling variables?

Originally Posted by Jake
Like I said earlier, I wouldn't.
Oke, So it is oke to just mention that I did a standard multiple regression analysis while controlling for the effects of gender, education and age?

For me this also more beneficial since I am not familiar with hierarchical regression analysis at all...

12. ## Re: hierarchical regression for controlling variables?

In my opinion, absolutely.

13. ## Re: hierarchical regression for controlling variables?

Seems like a good plan. It does not seem like hierarchical regression. I refereed a manuscript a couple of weeks ago and they said they used hierarchical regression and it seemed like what Jake described. My critique was to better describe the model and why it was hierarchical. Were fixed or random effects incorporated?

14. ## Re: hierarchical regression for controlling variables?

Thank you!

How should I interpret the model after including the control variables? Should I look to Block 2 of the table (with all variables included) and check the R squared value instead of the R square change value?

15. ## Re: hierarchical regression for controlling variables?

Originally Posted by hlsmith
Seems like a good plan. It does not seem like hierarchical regression. I refereed a manuscript a couple of weeks ago and they said they used hierarchical regression and it seemed like what Jake described. My critique was to better describe the model and why it was hierarchical. Were fixed or random effects incorporated?
I guess fixed but i am not sure